JonathanTonkin Ecologist

Research Interests

Research overview

I am a quantitative community ecologist aiming to disentangle the mechanisms that promote and maintain biodiversity at different scales. Much of my research involves searching for mechanistic links between environmental variation and population and community dynamics in highly dynamic, abiotically-forced systems. Although my research has evolved to be more computationally driven, my work maintains a strong empirical component, with considerable field experience in New Zealand, northern Spain, central China and western USA. This experience has typically centered on running waters, but my interests are question oriented, not system specific. I use a variety of approaches to explore these questions, including field manipulations, simulations and mining large datasets using different statistical, network, and machine learning techniques.

Much of my past and ongoing research focuses on the metacommunity ecology of dendritic ecological networks. The dendritic nature of river systems can promote unique metacommunity dynamics through regulating dispersal processes. Changes in connectivity and isolation along the network of these hierarchically organized systems can lead to unique patterns of biodiversity at all scales (alpha, beta, gamma). Understanding these physical constraints on metacommunity dynamics is not limited to basic questions, but also apply to key management issues. For instance, I have shown that for restoration projects to achieve their desired ecological outcome, spatial processes such as dispersal routes and barriers, and metacommunity dynamics need to be incorporated into planning. I intend to expand this recent work by applying metacommunity theory to stream networks as a model system, including both aquatic and terrestrial organisms with different dispersal modes that depend on river networks. However, a fundamental problem often faced in stream metacommunity research is that of poor explanatory power and predictive ability. I believe that one of the central elements contributing to this problem is the stochastic nature of rivers and the lack of attention to temporal dynamics. This calls for two advancements in this area of research, which I plan to focus on the coming years: 1. More advanced modeling techniques to account for this stochastic element. 2. More complete time-series data on metacommunities.

This focus on dynamic systems has led on to recent work exploring how temporal environmental variability (seasonality and predictability) allows for the coexistence of a great number of species through temporal niche differentiation. Currently, we seek to understand how climate change will alter such relationships in river systems. In searching for a mechanistic link between climate, streamflow and population dynamics, we are developing novel modeling techniques incorporating stochastic coupled structured population models. These models incorporate detailed stage-specific vital rates of organisms that respond to specific aspects of the hydrologic regime (floods, droughts, etc.). This “interaction neutral” modeling approach, through the omission of interactions during model parameterization, sits somewhere between neutral biodiversity models and more complete food-web models that require a priori pairwise interactions. A simple implementation of this model can be found at Using network theory, we have been expanding on this work to explore the community-wide consequences of flow modification on river ecosystems.

My overarching goal is to contribute strong advances to ecological theory through searching for mechanisms regulating biodiversity in dynamic abiotically-forced systems. As riverine systems are particularly stressed globally, much of my work has had a strong applied focus, including river regulation, mitigating migratory fish passage through novel solutions, and river restoration. By bringing together ecological theory with key environmental issues, I hope to provide tools for practitioners wishing to better manage threatened ecosystems.